Conference report

Standardisation of methods for assessing mould germination:A workshop report

Philippe Dantigny a,⁎, Maurice Bensoussan a, Valerie Vasseur b, Ahmed Lebrihi c, Claude Buchet d,Mustapha Ismaili-Alaoui e, Frank Devlieghere f, Sévastianos Roussos g

a Laboratoire de Microbiologie, UMR Université de Bourgogne/INRA 1232, ENS.BANA, 1 Esplanade Erasme, F-21000 Dijon, Franceb Ecole Supérieure de Microbiologie et Sécurité Alimentaire de Brest, Laboratoire de Biodiversité et Ecologie Microbienne, Technopole Brest-Iroise,

F-29280 Plouzane, Francec Laboratoire de Génie Chimique, UMR5503 (CNRS/INPT/UPS), Département Bioprocédés et Systèmes Microbiens, 1 Avenue de l'Agrobiopole, BP 32607,

F-31326 Castanet-Tolosan, Franced Degussa Ferments d'Aromatisation, 16 rue de la gare, F-77260 La Ferté-Sous-Jouarre, France

e Institut Agronomique et Vétérinaire Hassan II, BP 6202, Rabat, Moroccof Laboratory for Food Microbiology and Food Preservation, Department of Food Safety and Food Quality, Coupure Links 653, Ghent University,

B-9000 Ghent, Belgiumg Laboratoire de Microbiologie de l'IRD, ESIL, Case 925, Universités de Provence et de la Méditerranée, 163 Avenue de Luminy;

F-13288 Marseille cedex 9, France

Abstract

The first workshop on predictive mycology was held in Marseille, France, 2–4 February 2005 under the auspices of the FrenchMicrobiological Society. The purpose of the workshop was to list the different techniques and definitions used by scientists for assessingmould germination and to evaluate the influence of the different techniques on the experimental results. Recommendations were made whena large consensus was obtained. In order to facilitate the study of germination, alternative methods to microscopic examination wereexamined.

Keywords: Mould; Fungi; Germination; Standardisation; Methods

1. Introduction

Predictive modelling of filamentous fungi growth has notreceived the same level of attention as that of bacterial growth(Gibson and Hocking, 1997). Although this observation is true,to date the number of studies devoted to fungal developmenthad increased significantly. Fungal development involves ger-mination followed by growth. Whereas, growth is usuallyreported as the measurement of the radial growth rate in milli-meter per day, techniques and definitions for assessing mouldgermination vary greatly between authors. Therefore germina-tion data obtained in the literature are difficult to compare. This

⁎ Corresponding author. Tel.: +33 380396671; fax: +33 380396640.E-mail address: phdant@u-bourgogne.fr (P. Dantigny).

is of major concern because a lot of time is required forobtaining these experimental data.

Germination can be considered as the main step to be fo-cused on. In fact, visible mycelium appears shortly after ger-mination is completed. Germination requires microscopicobservation for evaluating the length of the germ tube (Dan-tigny et al., 2005a). Moreover, observations and measurementsshould be carried out without opening the dishes (Magan andLacey, 1984) and experimental devices should be developed forthis purpose. In order to provide reproducible, accurate data,there is an urgent need for standardising methods for assessingmould germination.

Spores, which are responsible for mould dissemination, arealways used to study germination. Therefore, the way that sporesare obtained, stored, prepared, inoculated and grownmay have aninfluence on the germination kinetics. Two important definitions

were discussed, these being when is a spore considered germi-nated and what is the most suitable definition of the germinationtime? In order to accurately determine biological parameters suchas the germination time, models should be used such as thoseapplied to determining bacterial growth rate. The type of modelchosen, as compared to another one, may affect the parametervalues. A model may also be based either implicitly or explicitly,on assumptions about the physiology of the germination. Even-tually, the key step of the modelling procedure is the validation ofthe model on food products and raw materials. Because theproducts are often opaque, it is clear that any germinationmodel will be difficult to validate. Alternative methods to micro-scopic observations should be developed.

Predictive mycology aims at predicting mould developmentand mycotoxin production in food and raw products (Dantigny,2004). For this purpose, models that take into account mouldspecificities are developed. However, this approach is not re-stricted to moulds responsible for food spoilage and mycotoxinproduction. As a primary step in mould development, germina-tion is also of great concern in biotechnology for producingstarters or fungal metabolites.

Every two years, a meeting dedicated to moulds is organisedby the French Microbiological Society. In 2005, two sessions ofthe 4th meeting were devoted to predictive mycology. Thesesessions were followed by a workshop on mould germination.The workshop was articulated into five major sections (i.e.,techniques, definitions, models, physiology of the germination,and alternative methods to microscopic observations).

About 30 scientists from different countries, working indifferent fields (either food mycology or biotechnology) andbelonging to academic institutions or industry, participated inthe workshop. They are studying different kind of moulds (i.e.,Zygomycetes, Ascomycetes and Deuteromycetes) that aregrown on different substrates under various environmental con-ditions. Due to the diversity of the participants, we believe thatthis workshop report may be useful to the whole community ofmycologists.

2. Techniques

2.1. Spore preparation

Fresh spores are obtained after mycelium was grown onvarious semi-synthetic media (aw 0.99) such as Potato DextroseAgar (PDA) or Malt Extract Agar (MEA). However, mediacharacterised by lower aw are required for growing xerophilicfungi such as Wallemia sebi. With the objective of producingfresh spores, the environmental conditions are usually set at theoptimum for growth in terms of water activity (aw) and temper-ature (T). Depending on the radial growth rate of the colony, theincubation time for obtaining conidia produced by most of theDeuteromycetes ranged from 5–12 days, but longer incubationtimes may be necessary for particular spores. For example, 19–98 days were necessary to produce macroconidia from Fusar-ium graminearum (Beyer et al., 2004) whereas 67±11 dayswere necessary to produce ascospores from Gibberella zeae,(teleomorph F. graminearum) Beyer et al. (2005).

The surface of aerial fungal cultures is flooded either withdistilled water, saline water or tryptone salt. Very often awetting agent, Tween80 at a concentration 0.01–0.05% (vol /vol), is added to the solution. The normalisation protocol NF V18–301 for determining the number of mould spores in cattlefood recommended 0.033% (vol /vol) Tween80 (AFNOR,1983). At these concentrations, the effect of the diluting solu-tion on the lag time for growth, that can be related under certainconditions to the germination time (see Section 6), was notsignificant (Sautour et al., 2003). Therefore no particular rec-ommendation was made on choice of suspending solution.

Whatever the use of the spore suspension, all participantsagreed that the initial time was the time at which spores weresuspended in the solution. In fact, spores can swell as soon asthey were suspended in an aqueous solution. As stated in thenext section, spore suspensions are usually standardised to acertain concentration. Therefore, it was recommended that di-lution of the spore suspensions, if necessary, be done quickly aspossible. This definition applies without any problem to freshspores, but may cause some concern with frozen or freeze-driedspores.

Some participants are using sterile bent glass rods for open-ing the sporangium of Zygomycetes, but the technique consist-ing of gently brushing the surface of the mycelium with aninoculating loop or other similar device is also suitable for thispurpose.

2.2. Inoculation protocols

Inoculating mature spores as a needlepoint is probably theclosest technique to what was happening naturally. Due to apossible alteration of the water activity by pouring a solution,this technique is used for studying xerophilic fungi (Pitt andHocking, 1977). This technique is suitable for studying mouldgrowth, but with some fungi, especially Penicillium and Asper-gillus, it is important to minimise colonies from stray spores(Pitt and Hocking, 1999). However this technique cannot bestandardised easily. Most of the recent germination studies wereusing standardised inoculum. Two groups can be defineddepending on whether spore suspensions were spread or poured.

2.2.1. Spreading techniquePetri dishes are swirled so that the inoculum covered the

entire surface of the agar. Generally, a high number of spores(greater than 1.0 104) and a large volume of suspension areutilised (see Table 1). When spore suspensions are spreadover the surface of the agar medium, some patches of thismedium are aseptically removed at regular intervals for mi-croscopic observations. Therefore, this technique is invasive,time-consuming and the risk of post-inoculation contamina-tion is increased. In addition, the same spore cannot bestudied throughout germination.

2.2.2. Pouring techniqueUsing this technique a small volume of suspension (less than

100 μl) is inoculated at the same spot (see Table 1). Therefore,fewer spores than those spread with the other technique are

Table 1Comparison between the spreading and the pouring inoculation techniques cited in the literature

Inoculation technique Volume Number of spores Moulds References

Spreading 10 μl 1.5 104 Aspergillus niger Plascencia-Jatomea et al. (2003)50 μl 5.0 104 Microsphaeropsis spp. Carisse and Bernier (2002)0.1 ml 1.0 106–5.0 106 Aspergillus ochraceus Pardo et al. (2005)0.1 ml 1.0 105 Penicillium digitatum, P. italicum,

Geotrichum candidumPlaza et al. (2003)

0.1 ml 1.0 105–5.0 105 Fusarium moniliforme, F. proliferatum Marín et al. (1996)0.5 ml 0.5–1.0 106 Coniothyrium minitans McQuilken et al. (1997)

Pouring 1 μl 225 Mucor racemosus Dantigny et al. (2002)5 μl 1.7 104–2.0 104 Wallemia sebi Patriarca et al. (2001)10 μl 1.0 104 Penicillium chrysogenum Sautour et al. (2001b)60 μl 300 Leptosphaeria maculans Huang et al. (2001)

inoculated. In contrast to previous technique, spores can beexamined microscopically without opening the dishes.

Although we do not recommend any particular technique asthey are both widely used. The final objective was to monitorfrom 10 to 20 spores per microscope field at 100×–400×.

2.3. Experimental device for monitoring germination

An experimental device developed for an easy assessment ofgermination was shown in Fig. 1.

The device aimed at observing microscopically the sporesthrough the lid, without opening the dishes. In each dish,three patches of germination medium were applied to thelid, and different control solutions can be poured into thebottom. To apply the medium to the lid, a sterile dish wasopened in a laminar flow cabinet. Three small glass cylinderswere placed on the internal side of the lid and filled withsterile germination medium to a thickness of about 1 mm.After the medium had solidified, the glass cylinders wereremoved and each surface was ready for inoculation. A drop-let of spore suspension was poured at the centre of eachpatch. After inoculation, Petri dishes were hermeticallyclosed. Providing Petri dishes were marked, the germinationof the same spore can be monitored throughout the experi-ments (Dantigny et al., 2005b).

3. Definitions

3.1. When is a spore germinated?

The definition of when a spore is considered germinated isbased on a comparison between the length of the germ tube andthe diameter of the spore that is assumed round shaped.

g

glass cylinder

Petri dish

water + gly

Fig. 1. Example of experimental device developed for observing mould sp

Depending on whether spores were considered germinated,(when the length of the germ tube was equal to (i) one half or(ii) twice the spore diameter), the germination time of Mucorracemosus varied by about 30% and 25% at 25 and 15 °C,respectively.

It was shown that the definition had an important effect onthe germination time. However, none of these definitions wasused. All participants agreed to consider a spore germinatedwhen the length of the longest germ tube was greater to equalthe greatest dimension of the swollen spore. This definition thatapplied also to a non-round shaped spore and to a spore withmore than one germ tube was also in accordance with theliterature.

The development of image analysis for monitoring sporegermination (Paul et al., 1993; Oh et al., 1996) was mentioned.However, few laboratories have developed this technique. Itwould have been unrealistic to provide a definition of a tech-nique that is not routinely used.

3.2. Germination time

Germinated spores eventually form mycelium. Based on thisobservation, the germination time was defined by some authorsas the time at which mycelium was observed (Table 2). But, inmost cases the germination time was based on microscopicobservations. A morphological definition was given for defin-ing a germinated spore in the Section 3.1. When a population ofspores is considered, the definition should be based on anothercriterion.

In fact, all spores were not germinated at the same time.Each spore was characterised by an individual germinationtime. For a population, the percentage of germinated sporeswas calculated as: P (%)=(Ngerminated spores /Ntotal spores)×100.

ermination medium

Parafilm®

cerol

lid

ores microscopically without opening the dish (Sautour et al., 2001a).

Table 2Definitions of the germination time reported in the literature

Definition Percentage criterion Reference

Change in theappearance ofthe inoculum

na Ayerst (1969)

Visible mycelium oninoculated points

1 /50th El Halouat and Debevere(1997)

Significant numberof germ tubes

na Pitt and Hocking (1977)

ns ns Hocking and Pitt (1979)ns ns Wheeler et al. (1988)Percentage of

germinated spores10% of total spores Magan and Lacey (1984)

Percentage ofgerminated spores

50% of total spores Hocking and Miscamble(1995)Patriarca et al. (2001)

Percentage ofgerminated spores

90% of total spores Sautour et al. (2001b)

Percentage ofgerminated spores

50% of viable spores Huang et al. (2001)

na: not applicable, ns : not stated.

From an initial value, P0=0%, the percentage of germinatedspores increased with time up to a maximum percentage ofgermination, Pmax. The definition of the germination timevaried greatly, depending on the source (Table 2). It appearedduring the workshop that people concerned with food spoil-age moulds were setting the germination time definition at alow percentage (say 10%). In contrast, people concerned withproducing starters or fungal metabolites were using a greaterpercentage (say 90%).

The germination time, tG, is dependent on the percentageset for its definition, especially under non-optimal conditions.For example, the germination time of Fusarium moniliformeat 0.90 aw was equal to 36 and to 80 h, at P=10% andP=90%, respectively (calculated from the data of Marín etal., 1996). The need for having a precise definition of thegermination time was illustrated in Table 2. Chronologicallythe definitions given by Hocking vary from a significantnumber of the germ tubes (Pitt and Hocking, 1977) to apercentage of germinated spores equal to 50% (Hockingand Miscamble, 1995).

It was decided during the workshop that 50% represented agood trade-off between 10% and 90%. In addition, this valuehas been widely used in the literature as seen in Table 2.However, under harsh environmental conditions some sporeswere unable to initiate a germ tube, thus leading to a maximumpercentage of germination less than 100%. For this reason, thedefinition based on a percentage of germinated spores equal to50% of the viable spores (Huang et al., 2001) should bepreferred.

4. Models

Biological growth parameters (e.g., radial growth rate, lagtime, germination time…) are very often calculated from modelequations. The probability that a time at which spores are exam-ined coincided with germination time is close to zero. Therefore,

spores are examined at regular intervals, thus enabling a germi-nation curve: P (%)= f(t) to be drawn. Two models for fitting thegermination data can be used (Dantigny et al., 2003):

the Gompertz equation:

P ¼ Adexpd −explmdeð1Þ

Aðkge−tÞ þ 1

� �� �ð1Þ

where A (%) was the asymptotic P value at t→+∝, μm (h−1) wasthe slope term of the tangent line through the inflection point (tG)as defined further, λge (h) was the t-axis intercept of the tangentthrough the inflection point and t was the time (h).

the logistic function: P ¼ Pmax

1þ exp½kðs−tÞ� ð2Þ

where Pmax (%) was the asymptotic P value at t→+∝, k (h−1)was the slope term of the tangent line through the inflection point,τ (h) was the inflection point where P equals half of the Pmax and twas the time (h).

By using any of these equations, it is possible to estimateaccurately the time necessary to reach a certain percentage ofgermination. For example, the time at which 50% of the max-imum percentage of germination is defined as:

tG ¼ kge þ A=ðlmeð1ÞÞ andtG ¼ s for equations ð1Þ and ð2Þ ; respectively:

Preliminary results on the germination of Penicillium chry-sogenum had demonstrated that the choice of the model had nosignificant effect on the estimation of the germination time.However, from a theoretical point of view, the models aredifferent. Both the models exhibited an inflection point at tG.But the logistic function is symmetrical with respect to thatpoint, not the Gompertz equation. By choosing a model,assumptions on the physiology of the germination are madeimplicitly.

5. Physiology of the germination

Assuming the definition of the germination time as the timeat which 50% of the viable spores had germinated, two groupscan be distinguished within the population. Individual sporescharacterised by individual germination times less than tG (firstgroup), and greater than tG (second group). If the growth of thegerm tube of the whole population is constant, and if theindividual germination times amongst each group are equally,normally distributed, then the germination curve should besymmetrical as compared to the inflection point. Preliminaryresults on the germination of P. chrysogenum suggested thatexperimental data were fit more accurately with the logisticfunction than with the Gompertz equation.

At present, the available studies showed a linear increase ofthe length of the germ tubes during the early germination phase.Unfortunately, in these studies, the rate of elongation was deter-mined from an average of the measured length of 20 (Snow,1949) and 5 to 6 (Trinci, 1971) germ tubes. In order to verify thisassumption, experiments allowing the measurement of the

length of individual spores throughout germination should becarried out.

6. Alternative methods to microscopic observation

The development of alternative methods to microscopicobservation would appear a significant breakthrough in thestudies of germination. Firstly, microscopic observations aretedious and time-consuming. Secondly, the opacity of productshas made model validation very difficult. To overcome thesedisadvantages, alternative methods aimed at substituting micro-scopic observations for macroscopic ones. For example, anarbitrary parameter “rejection time” (i.e., the time required toform a 2 mm diameter colony) was introduced by Horner andAnagnostopoulos (1973) to express the shelf life of jam. Therejection time was affected by the lag for growth and the radialgrowth rate, μ (mm d−1) of the colony. An arithmetic plot ofincrease in colony radius, r, against time was, after the lagphase, λ (d), initially linear, as found for fungi by Trinci(1969). A simple linear model with breakpoint was used forrepresenting the plot:

r ¼ lðt−kÞ ð3Þ

To calculate the lag, this linear section of the graph isextrapolated to a zero increase in diameter. The intercept onthe time axis is defined as the lag. In order to substitutemicroscopic observations for macroscopic ones, the relation-ship between the lag and the germination time was assessed forM. racemosus (Dantigny et al., 2002). In these studies, it wasshown that, for a number of inoculated spores arbitrarily set to225, the lag coincided with the completion of the germinationprocess (say P≥95%). It was shown for P. chrysogenum thatthe lag was dependent on the initial number of spores. This wasexplained by the ability of a large number of spores to formmore readily a visible colony than a fewer number of spores(Sautour et al., 2003).

The effect of the number of spores inoculated, N, on the lagwas described by the following relationship (González et al.,1987):

k ¼ k1−kdlogðNÞ ð4Þwhere λ1 was the theoretical maximum lag obtained for onespore.

In order that the lag for growth equals the germination timeas defined in Section 3.2., a greater number of spores than 225should be inoculated. Eq. (4) can be used to calculate thenumber of spores that should be inoculated for having the lagequals to the germination time. This alternative approach fordetermining macroscopically the germination time should nowbe calibrated for various moulds.

7. Conclusions and recommendations

The number of participants to this workshop proved thatstandardising techniques for studying germination was a greatconcern. This workshop allowed these techniques to be listed.

Fortunately, the use of different techniques has no influence onthe acquisition of accurate raw data. For example, the nature ofthe diluting solution has no significant effect on germinationkinetics.

Germination studies were often based on microscopic obser-vations. For this purpose, two types of techniques for theinoculum preparation were described. The first one consistedof spreading the suspension of spores whereas the second oneinvolved applying a droplet onto the surface of the agar medi-um. The second technique allows microscopic observationswithout opening the dishes. More spores are spread than thosethat are poured. Hence the number of spores per microscopicfield is independent from the technique chosen.

In contrast, it was demonstrated that biological parameterssuch as the germination time were greatly dependent from thedefinitions. It was clear that people involved in food spoilagepreferred definitions of the germination time based on a verylow level of germinated spores. In contrast, people whoseinterest was producing starters or fungal metabolites advocatedgermination of all the spores. Other definitions were reported inthe literature. This observation was helpful in reaching a con-clusion that the accepted definition of the germination timeshould not be purpose dependent. It was also acknowledgedthat a definition for a biological parameter was often related to amodel.

The Gompertz equation and the logistic function are avail-able for describing germination kinetics. The time for which thecurve exhibited an inflection point can be easily calculatedfrom the model equations. In addition that time has a biologicalmeaning. It represents the time at which 50% of the viablespores had germinated (Huang et al., 2001). Accordingly, thatdefinition was recommended for the germination time.

A model is characterised by two aspects: (i) a predictiveone and (ii) a descriptive one. The predictive aspect of amodel allowed results to be obtained without carrying outadditional experiments. For this purpose, a model should bevalidated on a food product by means of challenge tests. Dueto the opacity of products, alternative methods for assessinggermination should be developed. These methods are based onmacroscopic observations. The plot of the colony radius ver-sus time can be extrapolated to calculate a lag. It was shownthat the germination time can be substituted for the lag pro-viding the inoculum size was carefully controlled. This ap-proach appeared effective but should be calibrated for themoulds of concern.

Two models (i.e., the Gompertz model and the logisticfunction) can be used for calculating the germination time. Byusing a model it was underlined that assumptions on the phys-iology of the germination were implicitly assumed. A modelenables the codification of the knowledge and assumption ofthe physiology. The distribution of the rate of elongation of thegerm tubes amongst spores remains unknown, thus preventingfrom favouring a model.

Predictive mycology aiming at predicting fungal growth andproduction of metabolites is not restricted to food spoilage. Thisfield can extend its use to biotechnology, phytopathology, assuggested by the large audience of the workshop.

References

Agence Française de Normalisation (AFNOR), 1983. Aliments des animaux.Dénombrement des moisissures, pp. 18–301. NF V.

Ayerst, G., 1969. The effects of moisture and temperature on growth and sporegermination in some fungi. Journal of Stored Product Research 5, 127–141.

Beyer, M., Röding, S., Ludewig, A., Verreet, J.−A., 2004. Germination andsurvival of Fusarium graminearum macroconidia as affected by environ-mental factors. Journal of Phytopathology 152, 92–97.

Beyer, M., Verreet, J.−A., Ragab, W.S.M., 2005. Effect of relative humidity ongermination of ascospores and macroconidia of Gibberella zeae anddeoxynivalenol production. International Journal of Food Microbiology98, 233–240.

Carisse, O., Bernier, J., 2002. Effect of environmental factors on growth,pycnidial production and spore germination of Microsphaeropsis isolateswith biocontrol potential against apple scab. Mycological Research 106,1455–1462.

Dantigny, P., 2004. Predictive mycology. In: McKellar, R.C., Lu, X. (Eds.),Modeling Microbial Responses in Food. CRC Press, Boca Raton, FL,pp. 313–320.

Dantigny, P., Soares Mansur, C., Sautour, M., Tchobanov, I., Bensoussan, M.,2002. Relationship between spore germination kinetics and lag time duringgrowth ofMucor racemosus. Letters in Applied Microbiology 35, 395–398.

Dantigny, P., Guilmart, A., Bensoussan, M., 2003. Predictive mycology: somedefinitions. Cryptogamie. Mycologie 24, 377–383.

Dantigny, P., Guilmart, A., Bensoussan, M., 2005a. Basis of predictivemycology. International Journal of Food Microbiology 100, 187–196.

Dantigny, P., Tchobanov, I., Bensoussan, M., Zwietering, M.H., 2005b.Modelling the effect of ethanol vapour on the germination time ofPenicillium chrysogenum. Journal of Food Protection 68, 1203–1207.

El Halouat, A., Debevere, J.M., 1997. Effect of water activity; modifiedatmosphere packaging and storage temperature on spore germination ofmoulds isolated from prunes. International Journal of Food Microbiology35, 41–48.

Gibson, A.M., Hocking, A.D., 1997. Advances in the predictive modelling offungal growth in food. Trends in Food Science Technology 8, 353–358.

González, H.H.L., Resnik, S.L., Vaamonde, G., 1987. Influence of inoculumsize on growth rate and lag phase of fungi isolated from Argentine corn.International Journal of Food Microbiology 4, 111–117.

Hocking, A.D., Miscamble, B.F., 1995. Water relations of some Zygomycetesisolated from food. Mycological Research 99, 1113–1115.

Hocking, A.D., Pitt, J.I., 1979. Water relations of some Penicillium species at 25°C. Transactions of the British Mycological Society 73, 141–145.

Horner, K.J., Anagnostopoulos, G.D., 1973. Combined effects of water activity,pH and temperature on the growth and spoilage potential of fungi. Journal ofApplied Bacteriology 36, 427–436.

Huang, Y.J., Toscano-Underwood, C., Fitt, B.D.L., Todd, A.D., West, J.S.,Koopmann, B., Balesdent, M.H., 2001. Effects of temperature ongermination and hyphal growth from ascospores of A-group and B-groupLeptospaeria maculans (phoma stem cancer of oilseed rape). Annals ofApplied Biology 139, 193–207.

Magan, N., Lacey, J., 1984. Effect of temperature and pH on water relations offield and storage fungi. Transactions of the British Mycological Society 82,71–81.

Marín, S., Sanchis, V., Teixido, A., Saenz, R., Ramos, A.J., Vinas, I., Magan, N.,1996. Water and temperature relations and microconidial germination ofFusarium moniliforme and Fusarium proliferatum from maize. CanadianJournal of Microbiology 42, 1045–1050.

McQuilken, M.P., Budge, S.P., Whipps, J.M., 1997. Effects of culture media andenvironmental factors on conidial germination, pycnidial production andhyphal extension of Coniothyrium minitans. Mycological Research 101,11–17.

Oh, K.−B., Chen, Y., Matsuoka, H., Yamamoto, A., Kurata, H., 1996.Morphological recognition of fungal spore germination by a computer-aided image analysis and its application to antifungal activity evaluation.Journal of Biotechnology 45, 71–79.

Pardo, E., Ramos, A.J., Sanchis, V., Marín, S., 2005. Modelling of effects ofwater activity and temperature on germination and growth of ochratoxigenicisolates of Aspergillus ochraceus on a green coffee-based medium.International Journal of Food Microbiology 98, 1–9.

Patriarca, A., Vaamonde, G., Fernández Pinto, V., Comerio, R., 2001. Influenceof water activity on the growth of Wallemia sebi: application of a predictivemodel. International Journal of Food Microbiology 68, 61–67.

Paul, G.C., Kent, C., Thomas, C.R., 1993. Viability testing and characterisationof germination of fungal spores by automatic image analysis. Biotechnologyand Bioengineering 42, 11–23.

Pitt, J.I., Hocking, A.D., 1977. Influence of solute and hydrogen ionconcentration on the water relations of some xerophilic fungi. Journal ofGeneral Microbiology 101, 35–40.

Pitt, J.I., Hocking, A.D., 1999. Fungi and Food Spoilage. Aspen Publishers Inc.,Gaithersburg, MD. 593 pp.

Plascencia-Jatomea, M., Viniegra, G., Olayo, R., Castillo-Ortega, M.M., Shirai,K., 2003. Effect of chitosan and temperature on spore germination ofAspergillus niger. Macromolecular Bioscience 3, 582–586.

Plaza, P., Usall, J., Teixidó, N., Viňas, I., 2003. Effect of water activity andtemperature on germination and growth of Penicillium digitatum, P. italicumand Geotrichum candidum. Journal of Applied Microbiology 94, 549–554.

Sautour, M., Rouget, A., Dantigny, P., Divies, C., Bensoussan, M., 2001a.Prediction of conidial germination of Penicillium chrysogenum as influencedby temperature, water activity and pH. Letters in Applied Microbiology 32,131–134.

Sautour, M., Dantigny, P., Divies, C., Bensoussan, M., 2001b. Application ofDoehlert design to determine the combined effects of temperature, wateractivity and pH on conidial germination of Penicillium chrysogenum.Journal of Applied Microbiology 91, 900–906.

Sautour, M., Dantigny, P., Guilhem, M.-C., Bensoussan, M., 2003. Influence ofinoculum preparation on the growth of Penicillium chrysogenum. Journal ofApplied Microbiology 95, 1034–1038.

Snow, D., 1949. The germination of mould spores at controlled humidities.Annals of Applied Biology 36, 1–13.

Trinci, A.P.J., 1969. A kinetic study of the growth of Aspergillus nidulans andother fungi. Journal of General Microbiology 57, 11–24.

Trinci, A.P.J., 1971. Exponential growth of the germ tubes of fungal spores.Journal of General Microbiology 67, 345–348.

Wheeler, K.A., Hocking, A.D., Pitt, J.I., 1988. Water relations of someAspergillus species isolated from dried fish. Transactions of the BritishMycological Society 91, 631–637.